artificial intelligence
Page: 17
Universal Music Group general counsel/executive vp of business and legal affairs, Jeffery Harleston, spoke as a witness in a Senate Judiciary Committee hearing on AI and copyright on Wednesday (July 12) to represent the music industry. In his remarks, the executive called for a “federal right of publicity” — the state-by-state right that protects artists’ likenesses, names, and voices — as well as for “visibility into AI training data” and for “AI-generated content to be labeled as such.”
Harleston was joined by other witnesses including Karla Ortiz, a conceptual artist and illustrator who is waging a class action lawsuit against Stability AI; Matthew Sag, professor of artificial intelligence at Emory University School of Law; Dana Rao, executive vp/general counsel at Adobe; and Ben Brooks, head of public policy at Stability AI.
“I’d like to make four key points to you today,” Harleston began. “First, copyright, artists, and human creativity must be protected. Art and human creativity are central to our identity.” He clarified that AI is not necessarily always an enemy to artists, and can be used in “service” to them as well. “If I leave you with one message today, it is this: AI in the service of artists and creativity can be a very, very good thing. But AI that uses, or, worse yet, appropriates the work of these artists and creators and their creative expression, their name, their image, their likeness, their voice, without authorization, without consent, simply is not a good thing,” he said.
Second, he noted the challenges that generative AI poses to copyright. In written testimony, he noted the concern of “AI-generated music being used to generate fraudulent plays on streaming services, siphoning income from human creators.” And while testifying at the hearing, he added, “At Universal, we are the stewards of tens of thousands, if not hundreds of thousands, of copyrighted creative works from our songwriters and artists, and they’ve entrusted us to honor, value and protect them. Today, they are being used to train generative AI systems without authorization. This irresponsible AI is violative of copyright law and completely unnecessary.”
Training is one of the most contentious areas of generative AI for the music industry. In order to get an AI model to learn how to generate a human voice, a drum beat or lyrics, the AI model will train itself on up to billions of data points. Often this data contains copyrighted material, like sound recordings, without the owner’s knowledge or compensation. And while many believe this should be considered a form of copyright infringement, the legality of using copyrighted works as training data is still being determined in the United States and other countries.
The topic is also the source of Ortiz’s class action lawsuit against Stability AI. Her complaint, filed in California federal court along with two other visual artists, alleges that the “new” images generated by Stability AI’s Stable Diffusion model used their art “without the consent of the artists and without compensating any of those artists,” which they feel makes any resulting generation from the AI model a “derivative work.”
In his spoken testimony, Harleston pointed to today’s “robust digital marketplace” — including social media sites, apps and more — in which “thousands of responsible companies properly obtained the rights they need to operate. There is no reason that the same rules should not apply equally to AI companies.”
Third, he reiterated that “AI can be used responsibly…just like other technologies before.” Among his examples of positive uses of AI, he pointed to Lee Hyun [aka MIDNATT], a K-pop artist distributed by UMG who used generative AI to simultaneously release the same single in six languages using his voice on the same day. “The generative AI tool extended the artist’s creative intent and expression with his consent to new markets and fans instantly,” Harleston said. “In this case, consent is the key,” he continued, echoing Ortiz’s complaint.
While making his final point, Harleston urged Congress to act in several ways — including by enacting a federal right of publicity. Currently, rights of publicity vary widely state by state, and many states’ versions include limitations, including less protection for some artists after their deaths.
The shortcomings of this state-by-state system were highlighted when an anonymous internet user called Ghostwriter posted a song — apparently using AI to mimic the voices of Drake and The Weeknd –called “Heart On My Sleeve.” The track’s uncanny rendering of the two major stars immediately went viral, urging the music business to confront the new, fast-developing concern of AI voice impersonation.
A month later, sources told Billboard that the three major label groups — UMG, Warner Music Group and Sony Music — have been in talks with the big music streaming services to allow them to cite “right of publicity” violations as a reason to take down songs with AI vocals. Removing songs based on right of publicity violations is not required by law, so the streamers’ reception to the idea appears to be voluntary.
“Deep fakes, and/or unauthorized recordings or visuals of artists generated by AI, can lead to consumer confusion, unfair competition against the artists that actually were the original creator, market dilution and damage to the artists’ reputation or potentially irreparably harming their career. An artist’s voice is often the most valuable part of their livelihood and public persona. And to steal it, no matter the means, is wrong,” said Harleston.
In his written testimony, Harleston went deeper, stating UMG’s position that “AI generated, mimicked vocals trained on vocal recordings from our copyrighted recordings go beyond Right of Publicity violations… copyright law has clearly been violated.” Many AI voice uses circulating the internet involve users mashing up one previously released song topped with a different artist’s voice. These types of uses, Harleston wrote, mean “there are likely multiple infringements occurring.”
Harleston added that “visibility into AI training data is also needed. If the data on AI training is not transparent, the potential for a healthy marketplace will be stymied as information on infringing content will be largely inaccessible to individual creators.”
Another witness at the hearing raised the idea of an “opt-out” system so that artists who do not wish to be part of an AI’s training data set will have the option of removing themselves. Already, Spawning, a music-tech start-up, has launched a website to put this possible remedy into practice for visual art. Called “HaveIBeenTrained.com,’ the service helps creators opt-out of training data sets commonly used by an array of AI companies, including Stability AI, which previously agreed to honor the HaveIBeenTrained.com opt-outs.
Harleston, however, said he did not believe opt-outs are enough. “It will be hard to opt out if you don’t know what’s been opted in,” he said. Spawning co-founder Mat Dryhurst previously told Billboard that HaveIBeenTrained.com is working on an opt-in tool, though this product has yet to be released.
Finally, Harleston urged Congress to label AI-generated content. “Consumers deserve to know exactly what they’re getting,” he said.
From ChatGPT writing code for software engineers to Bing’s search engine sliding in place of your bi-weekly Hinge binge, we’ve become obsessed with the capacity for artificial intelligence to replace us.
Within creative industries, this fixation manifests in generative AI. With models like DALL-E generating images from text prompts, the popularity of generative AI challenges how we understand the integrity of the creative process: When generative models are capable of materializing ideas, if not generating their own, where does that leave artists?
Google’s new text-based music generative AI, MusicLM, offers an interesting answer to this viral terminator-meets-ex-machina narrative. As a model that produces “high-fidelity music from text descriptions,” MusicLM embraces moments lost in translation that encourages creative exploration. It sets itself apart from other music generation models like Jukedeck and MuseNet by inviting users to verbalize their original ideas rather than toggle with existing music samples.
Describing how you feel is hard
AI in music is not new. But between recommending songs for Spotify’s Discover Weekly playlists to composing royalty free music with Jukedeck, applications of AI in music have evaded the long-standing challenge of directly mapping words to music.
This is because, as a form of expression on its own, music resonates differently to each listener. The same way that different languages struggle to perfectly communicate nuances of respective cultures, it is difficult (if not impossible) to exhaustively capture all dimensions of music in words.
MusicLM takes on this challenge by generating audio clips from descriptions like “a calming violin melody backed by a distorted guitar riff,” even accounting for less tangible inputs like “hypnotic and trance-like.” It approaches this thorny question of music categorization with a refreshing sense of self awareness. Rather than focusing on lofty notions of style, MusicLM grounds itself in more tangible attributes of music with tags such as “snappy”, or “amateurish.” It broadly considers where an audio clip may come from (eg. “Youtube Tutorial”), the general emotional responses it may conjure (eg. “madly in love”), while integrating more widely accepted concepts of genre and compositional technique.
What you expect is (not) what you get
Piling onto this theoretical question of music classification is the more practical shortage of training data. Unlike its creative counterparts (e.g. DALL-E), there isn’t an abundance of text-to-audio captions readily available.
MusicLM was trained by a library of 5,521 music samples captioned by musicians called ‘MusicCaps.’ Bound by the very human limitation of capacity and the almost-philosophical matter of style, MusicCaps offers finite granularity in its semantic interpretation of musical characteristics. The result is occasional gaps between user inputs and generated outputs: the “happy, energetic” tune you asked for may not turn out as you expect.
However, when asked about this discrepancy, MusicLM researcher Chris Donahue and research software engineer Andrea Agostinelli celebrate the human element of the model. They describe primary applications such as “[exploring] ideas more efficiently [or overcoming] writer’s block,” quick to note that MusicLM does offer multiple interpretations of the same prompt — so if one generated track fails to meet your expectations, another might.
“This [disconnect] is a big research direction for us, there isn’t a single answer,” Andrea admits. Chris attributes this disconnect to the “abstract relationship between music and text” insisting that “how we react to music is [even more] loosely defined.”
In a way — by fostering an exchange that welcomes moments lost in translation — MusicLM’s language-based structure positions the model as a sounding board: as you prompt the model with a vague idea, the generation of approximates help you figure out what you actually want to make.
Beauty is in breaking things
With their experience producing Chain Tripping (2019) — a Grammy-nominated album entirely made with MusicVAE (another music generative AI developed by Google) — the band YACHT chimes in on MusicLM’s future in music production. “As long as it can be broken apart a little bit and tinkered with, I think there’s great potential,” says frontwoman Claire L. Evans.
To YACHT, generative AI exists as a means to an end, rather than the end in itself. “You never make exactly what you set out to make,” says founding member Jona Bechtolt, describing the mechanics of a studio session. “It’s because there’s this imperfect conduit that is you” Claire adds, attributing the alluring and evocative process of producing music to the serendipitous disconnect that occurs when artists put pen to paper.
The band describes how the misalignment of user inputs and generated work inspires creativity through iteration. “There is a discursive quality to [MusicLM]… it’s giving you feedback… I think it’s the surreal feeling of seeing something in the mirror, like a funhouse mirror,” says Claire. “A computer accent,” band member Rob Kieswetter jokes, referencing a documentary about the band’s experience making Chain Tripping.
However, in discussing the implications of this move to text-to-audio generation, Claire cautions the rise of taxonomization in music: “imperfect semantic elements are great, it’s the precise ones that we should worry about… [labels] create boundaries to discovery and creation that don’t need to exist… everyone’s conditioned to think about music as this salad of hyper-specific genre references [that can be used] to conjure a new song.”
Nonetheless, both YACHT and the MusicLM team agrees that MusicLM — as it currently is — holds promise. “Either way there’s going to be a whole new slew of artists fine-tuning this tool to their needs,” Rob contends.
Engineer Andrea recalls instances where creative tools weren’t popularized for its intended purpose: “the synthesizer eventually opened up a huge wave of new genres and ways of expression. [It unlocked] new ways to express music, even for people who are not ‘musicians.’” “Historically, it has been pretty difficult to predict how each piece of music technology will play out,” researcher Chris concludes.
Happy accidents, reinvention, and self-discovery
Back to the stubborn, unforgiving question: Will generative AI replace musicians? Perhaps not.
The relationship between artists and AI is not a linear one. While it’s appealing to prescribe an intricate and carefully intentional system of collaboration between artists and AI, as of right now, the process of using AI in producing art resembles more of a friendly game of trial and error.
In music, AI gives room for us to explore the latent spaces between what we describe and what we really mean. It materializes ideas in a way that helps shape creative direction. By outlining these acute moments lost in translation, tools like MusicLM sets us up to produce what actually ends up making it to the stage… or your Discover Weekly.
Tiffany Ng is an art & tech writer based in NYC. Her work has been published in i-D Vice, Vogue, South China Morning Post, and Highsnobiety.
R&B singer 6LACK has partnered with Endel to create alternate versions of his latest album Since I Have A Lover that is specially formulated to promote “restorative rest and mental balance” as part of BIPOC Mental Health Awareness Month. One alternate version is designed to promote sleep, out July 8, and the second will promote focus, out July 24.
To promote the collaboration, 6lack will host a live pre-listening session on the Endel app on July 6, and two in-person events will take place in LA and Berlin that same day.
Endel is a start-up that creates what it calls “functional sound,” a form of ambient music that supports listeners day-to-day wellness needs, including sleep, meditation and focus. While Endel’s proprietary music-making technology is powered by artificial intelligence, Endel does not create new songs out of thin air. Instead, it generates ambient soundscapes by rearranging pieces of music provided by artists in ways that Endel says promotes specific wellness goals.
6lack joins the likes of other artists like Grimes and James Blake who have partnered with the soundscape start-up in the past, however, the alternate version of Since I Have A Lover is a new kind of partnership for the wellness brand because it is the first full album Endel has remixed and released on streaming services.
In March, Oleg Stavitsky, co-founder and CEO of Endel, told Billboard he felt this was a way for the soundscape company to help major labels evolve. “We can process the stems [the audio building blocks of a track] from Miles Davis’ Kind of Blue and come back with a functional sleep version of that album,” Stavitsky said, adding that Endel was in talks with all the major labels about trying this at the time.
Since then, Endel has partnered with UMG to do just that, and Since I Have A Lover marks the companies’ first attempt. Their partnership came as a surprise to some, given UMG chief Lucian Grainge‘s negative recent remarks about AI and functional music. Lamenting that functional music drives streaming dollars away from pop music and towards rain sounds, white noise, and other ambient recordings, he said in a memo to staff in January that “consumers are increasingly being guided by algorithms to lower-quality functional content that in some cases can barely pass for ‘music.’”
But at the time the partnership was announced, Michael Nash, UMG evp and chief digital officer, praised Endel for “utiliz[ing] their patented AI technology” to create ambient music because it is “anchored in our artist-centric philosophy” and “powered by AI that respects artists’ rights in its development.”
Endel’s ambient soundscapes have been on streaming services since 2019, thanks to UMG competitor Warner Music Group. Known to be bullish in its investment and partnership strategy with emerging music tech companies, WMG signed Endel to a 20-album distribution deal.
“This is about letting people experience my music in a new way,” 6LACK says of his collaboration with Endel. “These sounds can be for rest and relaxation, or for helping you feel inspired and creative. It’s for finding a sense of balance in life. Since I Have a Lover has plenty of magical sounds, and combined with Endel’s AI and science, it was easy to create something that felt healing.”
“Using AI to reimagine your favorite music as a functional soundscape, designed to help solve the biggest mental health challenges we’re facing as a species, is our mission. 6LACK’s openness to experimentation and his ability to let go and trust the process was crucial to the success of this project,” says Stavitsky. “We’re extremely proud of the result and can’t wait for millions of people to experience the healing power of these soundscapes.”
Sony Music has created a new position of executive vp of AI, and hired former BPI CEO Geoff Taylor to take up the role, according to an internal memo obtained by Billboard. Taylor, who spent 15 years as chief of the British recorded music trade body, will report to Sony Music Entertainment COO Kevin Kelleher.
In the new role, Taylor will coordinate the major label’s business efforts surrounding artificial intelligence, and coordinate across the global digital business and business and legal affairs division, according to the memo.
The move comes as the music business continues to grapple with the particular challenges that are beginning to arise through the proliferation of artificial intelligence in the digital world, particularly its effect on copyright and ability to be trained on existing musical compositions, among other issues. Battles have begun to pop up around AI-generated music on streaming services, and songs that have been released that mimic existing artists, producers and songwriters without their input or consent, with additional use cases popping up seemingly every day.
The position appears to be the first AI-specific executive-level role introduced by any of the major labels. Check out Kelleher’s memo below.
All,
Artificial Intelligence (AI) has great significance for the future of the music industry and, as a result, more focused attention on it is required.
Accordingly, we are delighted to share that Geoff Taylor will be joining us as our new Executive Vice President, Artificial Intelligence.
Reporting to me and working closely with our Global Digital Business and Business & Legal Affairs divisions, Geoff will align and help coordinate the work of every part of the business that touches AI.
Geoff brings to our company decades of music industry experience. Most recently, from 2007 to 2022 Geoff was the Chief Executive of the BPI, our UK Trade Body for recorded music, where he led the fight against piracy and fraud and advocated for the strategic importance of recorded music to jobs, investment and maintaining the UK’s global competitiveness. Prior to joining the BPI, Geoff was General Counsel and Executive Vice-President at our global recorded music trade body, the IFPI from 2005-2007.
In these roles, Geoff has worked with our company for several years and I am delighted he is joining to help us successfully navigate a key moment in the history of the music industry.
So please join me in welcoming Geoff to Sony Music and feel free to reach out to him with any questions you might have at [Sony Music Entertainment Email Redacted].
Kevin
Lawyers for the RIAA are aiming to shut down a popular Discord server centered on artificial intelligence and voice models, the latest effort by music companies to rein in the disruptive new technology.
In an action filed last week in D.C. federal court, attorneys for RIAA obtained a subpoena demanding that Discord reveal the identities of users on “AI Hub,” a message board with 145,000 members that calls itself “a community dedicated to making AI voices and songs.”
In a letter to Discord presenting the company with the subpoena, the RIAA said those users had “infringed … copyrighted sound recordings” and that the tech company was required to hand over names, physical addresses, payment info, IP addresses and other identifying details.
The group’s lawyers also sent Digital Millennium Copyright Act takedown notices to Discord, first in late May and then again next week. The group demanded that Discord disable access to the server, remove or disable the infringing material, and inform the server’s users “of the illegality of their conduct.”
“This server [is] dedicated to infringing our members’ copyrighted sound recordings by offering, selling, linking to, hosting, streaming, and/or distributing files containing our members’ sound recordings without authorization,” the RIAA’s lawyers wrote in their June letter to Discord, which was obtained by Billboard. “We are asking for your immediate assistance in stopping this unauthorized activity.”
The subpoena against Discord was obtained under the DMCA’s Section 512(h), which enables rights holders like the RIAA’s members to unmask the identities of anonymous online infringers in certain circumstances.
Discord can fight back by seeking to “quash” the subpoena; Twitter won such a challenge last year, when a federal judge ruled that the First Amendment rights of a user trumped the need for an unmasking order. It could also refuse to honor the takedown, but that would put the site itself at risk of litigation.
As of Thursday evening (June 22), the main AI Hub server remained up on Discord; it was unclear if individual content or sub-channels had been removed. A spokesperson for the company did not return a request for comment.
In a statement to Billboard, an RIAA spokesperson confirmed that the group had taken the action against AI Hub. “When those who seek to profit from AI train their systems on unauthorized content, it undermines the entire music ecosystem – harming creators, fans, and responsible developers alike. This action seeks to help ensure that lawless systems that exploit the life’s work of artists without consent cannot and do not become the future of AI.”
The RIAA’s actions are just the latest sign that the explosive growth of AI technologies over the past year has sparked serious concerns in the music industry.
One big fear is that copyrighted songs are being used en masse to “train” AI models, all without any compensation going to the songwriters or artists that created them. In April, Universal Music Group demanded that Spotify and other streaming services prevent AI companies from doing so on their platforms, warning that it “will not hesitate to take steps to protect our rights.”
Another fear is the proliferation of so-called deepfake versions of popular music, like the AI-generated fake Drake and The Weeknd track that went viral in April. That song was quickly pulled down, but its uncanny vocals and mass popularity sparked concerns about future celebrity rip offs.
For RIAA, AI Hub likely triggered both of those worries. The server features numerous “voice models” that mimic the voices of specific real singers, including Michael Jackson and Frank Sinatra. And in the wake of the RIAA’s actions, users on the Discord server speculated Thursday that the takedowns were filed because users had disclosed that some of the models had been trained on copyrighted songs.
“We have had certain threats from record labels to takedown models, mainly because some posters decided to share datasets full of copyrighted music publicly,” one AI Hub admin wrote. “If you want to avoid unnecessary takedowns[,] most importantly, do NOT share the full dataset if you have copyrighted material in the dataset. The voice model itself is fine, but don’t share the dataset.”
As the music industry continues to grapple with the popularity and advancement of AI, some find themselves trying to figure out how to ethically incorporate artificial intelligence into their work. Earlier this month, when music titan Paul McCartney shared that he would be implementing AI in this process of finishing a new as-yet-untitled Beatles song, […]
Calling the rapid growth of artificial intelligence tools a “moment of revolution,” Senate Majority Leader Chuck Schumer said Wednesday that the government must act quickly to regulate companies that are developing it.
The New York Democrat said he is working on what he calls “exceedingly ambitious” bipartisan legislation to maximize the technology’s benefits and mitigate significant risks.
While Schumer did not lay out details of such legislation, he offered some key goals: protect U.S. elections from AI-generated misinformation or interference, shield U.S. workers and intellectual property, prevent exploitation by AI algorithms and create new guardrails to ward off bad actors.
AI legislation also should promote American innovation, Schumer said in a speech at the Center for Strategic and International Studies, a Washington think tank.
“If applied correctly, AI promises to transform life on Earth for the better,” Schumer said. “It will reshape how we fight disease, tackle hunger, manage our lives, enrich our minds and ensure peace. But there are real dangers that present themselves as well: job displacement, misinformation, a new age of weaponry and the risk of being unable to manage this new technology altogether.”
Schumer’s declaration of urgency comes weeks after scientists and tech industry leaders, including high-level executives at Microsoft and Google, issued a warning about the perils that artificial intelligence could pose to humankind.
“Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war,” their statement said.
Worries about artificial intelligence systems outsmarting humans and running wild have intensified in recent months with the rise of a new generation of highly capable AI chatbots such as ChatGPT. It has sent countries around the world scrambling to come up with regulations for the developing technology, with the European Union blazing the trail with its AI Act expected to be approved later this year.
On Tuesday, President Joe Biden convened a group of technology leaders in San Francisco to debate what he called the “risks and enormous promises” of artificial intelligence. In May, the administration brought together tech CEOs at the White House to discuss these issues, with the Democratic president telling them, “What you’re doing has enormous potential and enormous danger.”
“We’ll see more technological change in the next 10 years that we saw in the last 50 years,” Biden said.
White House chief of staff Jeff Zients’ office is developing a set of actions the federal government can take over the coming weeks regarding AI, according to the White House.
Schumer’s hands-on involvement in crafting AI legislation is unusual, as Senate leaders usually leave the task to individual senators or committees. But he has taken a personal interest in regulating the development of artificial intelligence, arguing that it is urgent as companies have already introduced human-like chatbots and other products that could alter life as we know it. He is working with another Democrat, Sen. Martin Heinrich of New Mexico, and Republican Sens. Mike Rounds of South Dakota and Todd Young of Indiana to speak with experts, educate colleagues and write the legislation.
It’s an unexpected role for Schumer, in particular, who famously carries a low-tech flip phone, and for the Senate as a whole, where the pace of legislation is often glacial.
Senators average around retirement age and aren’t known for their mastery of high-tech. They’ve been mocked in recent years for basic questions at hearings — asking Facebook founder Mark Zuckerberg simple questions about how his platform works at a 2018 hearing on Russian interference, for example — and for a bipartisan reluctance to regulate the technology industry at all.
Schumer, along with several Republican colleagues, say the federal government can no longer afford to be laissez-faire with tech companies.
“If the government doesn’t step in, who will fill its place?” Schumer asked. “Individuals and the private sector can’t do the work of protecting our country. Even if many developers have good intentions, there will always be rogue actors, unscrupulous companies, and foreign adversaries that seek to harm us. And companies may not be willing to insert guardrails on their own, certainly if their competitors are not required to insert them as well.”
Attempting to regulate AI, Schumer said, “is unlike anything Congress has dealt with before.”
It is unclear if Schumer will be able to accomplish his goals. The effort is in its earliest stages, with the bipartisan working group just starting a series of briefings for all 100 senators to get them up to speed. In the House, legislation to regulate or oversee artificial intelligence has been more scattershot, and Republican leaders have not laid out any ambitious goals.
Schumer acknowledged that there are more questions than answers about the technology.
“It’s not like labor or healthcare or defense where Congress has a long history we can work off of,” Schumer said. “In fact, experts admit nobody is even sure which questions policymakers should be asking. In many ways, we’re starting from scratch.”
The first time Ashley Elzinga, a 33-year-old DJ from Traverse City, Mich., heard her doppelgänger’s voice, she was not happy. Not because the sound of an artificial-intelligence imposter was so eerie. Not because AI technology portends robots might someday replace her. She didn’t like the way AI Ashley pronounced “news.” “I was like, ‘Why are they saying nooooose?’” recalls Elzinga in her flat Midwestern accent. “I was so embarrassed.”
It took a few tries for the engineers at Futuri Media, a Cleveland-based AI specialist, to find the right vocal balance between flat and sharp, deadpan and excited. “Now she’s ironing out…,” says Elzinga, a midday host for Portland Top 40 station Live 95.5, then corrects herself: “Now it’s ironing out…. She, or it, is starting to have more emotion and be a bit more accurate resemblance.”
AI Ashley, as Live 95.5 refers to the cloned voice on the air, made “her” debut on the air last Friday, delivering news, introducing songs and hyping station promotions in alternating speaking segments with the real Ashley Z, as Elzinga is known. Live 95.5 hasn’t received any listener complaints, says Dylan Salisbury, the station’s content director: “I don’t even know if they realize it yet.”
Alpha Media, owner of Live 95.5, started experimenting with AI voice technology last fall, according to Phil Becker, the company’s executive vp of content. When company execs learned Elzinga was about to take the full-time job in Traverse City, potentially reducing her hours on Live 95.5, they saw her as a “perfect storm” case study for an on-air test, he says: “The line in Moneyball is ‘the first guy through the wall always gets bloodied.’ That’s where we are right now. We’re OK playing some Moneyball-style radio, because it wins championships.”
Elzinga and Salisbury see AI as an efficiency tool, a way of stretching DJs’ hours so listeners can hear their voices even when they’re not physically present. For Elzinga, who multi-tasks her way through a full-time morning-show gig at her hometown Top 40 station WKHQ, then “tracks” her voice remotely for Live 95.5 and another station in Seattle every day, AI Ashley allows her to work even more. She owns the rights to her voice, approves every on-air AI usage and, Salisbury says, “We have increased her fee.”
“If she says stop, we have to stop,” Salisbury adds. “We’re trying to be respectful during the wild West of AI and go where we think the law is going to go.”
We made history as the world’s first radio station with an AI DJ! Our midday host Ashley has become AI Ashley! We can’t wait for you to meet Ashley, the world’s first artificially intelligent DJ. As to the intelligence of our other DJ’s…we’ll save that for another post 😉 pic.twitter.com/CtlMhYU0IO
— Live 95.5 (@live955) June 13, 2023
Of course, what is a neat, little, high-tech, mostly risk-free magic trick for Elzinga, Salisbury and Alpha Media, the Portland broadcast company that owns Live 95.5 and 205 other stations, is a terrifying prospect for much of the radio industry. When the station posted excitedly about AI Ashley last week, Twitter erupted: An NPR host tweeted an “everyone disliked that” meme, a freelance writer wanted to know, “Why would you participate in the very public elimination of your job?” and even J. Smith-Cameron, who plays Gerri on HBO’s Succession, wondered if Elzinga was “worried you’ll have ALL the days off now that they cloned you?”
For the past three decades, the broadcast industry has faced consolidation and extreme cost-cutting that has oftentimes meant layoffs of on-air talent. Over the past few years, DJs for local radio shows have been outsourced from other markets — much like Elzinga does in Portland and Seattle from her home in Michigan.
“They are eagerly stripping away, as fast as they can, the thing that makes radio unique,” says former radio host and station manager Michele Coppola, who’s now a Portland copywriter.
“My fear is there will be some owners that will [say], ‘This is an efficiency, this is a way for us to save money — that will further deplete the body of truly talented radio people,” adds Steve Reynolds, a Raleigh, N.C., talent coach for radio personalities.
“Futuri claims it’s going to be a tool, just like any other tool, to make a job easier,” says Lance Venta, owner and publisher of Radio Insight. “Voice-tracking, when used properly, is a tool. When it’s used to have a talent voice 35 stations to save money, it’s no longer a tool — it’s a weapon.”
Radio Waits
So far, the rest of the U.S. broadcast industry has yet to plunge into on-air AI voices as aggressively as Live 95.5. But radio stations around the world, and their digital competitors, have tinkered with the technology – and have suggested they may expand. In April, a Swiss station used AI to clone five human presenters on the air; comedian Drew Carey used an AI-generated approximation of his voice on his SiriusXM show in March; and in February, Spotify launched a (voiceless) AI-curated, personalized broadcast called “DJ.” During an April conference call about a soft advertising market, Bob Pittman, chairman and CEO of top broadcaster iHeartMedia, told investors after a 3.8% drop in revenue, “We and every other company are looking at how to use AI. I think AI can fundamentally change the cost structure of the company.”
At Audacy, the second-biggest broadcaster, execs have done a “fair bit of experimentation” with AI tools, from voice applications to ChatGPT-style generative text that helps produce drafts of advertising scripts, according to Jeff Sottolano, executive vp and head of programming. But he’s not convinced an AI Ashley-style experiment has “value it creates for the consumer,” because Alpha Media had to expend “up-front investment” on training, reviewing, post-production and editing — all of which, at least for now, contradict the company’s efforts for greater efficiency and cost-cutting.
“All that said, I expect it will continue to get better and easier and faster,” he says. “We don’t look at this as something that’s finished, but something that’s going to continue to evolve. Just because we haven’t done it today doesn’t mean we might not do it tomorrow.”
The human Ashley is happy with the AI arrangement as long as she and her robot counterpart are clearly identified as “Ashley Z” or “AI Ashley” every time she — or it — appears on the air. “You just need to make sure integrity comes first,” she says.
LONDON — Amid increasing concern among artists, songwriters, record labels and publishers over the impact of artificial intelligence (AI) on the music industry, European regulators are finalizing sweeping new laws that will help determine what AI companies can and cannot do with copyrighted music works.
On Wednesday (June 14), Members of the European Parliament (MEPs) voted overwhelmingly in favor of the Artificial Intelligence (AI) Act with 499 votes for, 28 against and 93 abstentions. The draft legislation, which was first proposed in April 2021 and covers a wide range of AI applications, including its use in the music industry, will now go before the European Parliament, European Commission and the European Council for review and possible amendments ahead of its planned adoption by the end of the year.
For music rightsholders, the European Union’s (EU) AI Act is the world’s first legal framework for regulating AI technology in the record business and comes as other countries, including the United States, China and the United Kingdom, explore their own paths to policing the rapidly evolving AI sector.
The EU proposals state that generative AI systems will be forced to disclose any content that they produce which is AI-generated — helping distinguish deep-fake content from the real thing — and provide detailed publicly available summaries of any copyright-protected music or data that they have used for training purposes.
“The AI Act will set the tone worldwide in the development and governance of artificial intelligence,” MEP and co-rapporteur Dragos Tudorache said following Wednesday’s vote. The EU legislation would ensure that AI technology “evolves and is used in accordance with the European values of democracy, fundamental rights, and the rule of law,” he added.
The EU’s AI Act arrives as the music business is urgently trying to respond to recent advances in the technology. The issue came to a head in April with the release of “Heart on My Sleeve,” the now-infamous song uploaded to TikTok that is said to have been created using AI to imitate vocals from Drake and The Weeknd. The song was quickly pulled from streaming services following a request from Universal Music Group, which represents both artists, but not before it had racked up hundreds of thousands of streams.
A few days before “Heart on My Sleeve” become a short-lived viral hit, UMG wrote to streaming services, including Spotify and Apple Music, asking them to stop AI companies from accessing the label’s copyrighted songs “without obtaining the required consents” to “train” their machines. The Recording Industry Association of America (RIAA) has also warned against AI companies violating copyrights by using existing music to generate new tunes.
If the EU’s AI Act passes in its present draft form, it will strengthen supplementary protections against the unlawful use of music in training AI systems. Existing European laws dealing with text and data-mining copyright exceptions mean that rightsholders will still technically need to opt out of those exceptions if they want to ensure their music is not used by AI companies that are either operating or accessible in the European Union.
The AI Act would not undo or change any of the copyright protections currently provided under EU law, including the Copyright Directive, which came into force in 2019 and effectively ended safe harbor provisions for digital platforms in Europe.
That means that if an AI company were to use copyright-protected songs for training purposes — and publicly declare the material it had used as required by the AI Act — it would still be subject to infringement claims for any AI-generated content it then tried to commercially release, including infringement of the copyright, legal, personality and data rights of artists and rightsholders.
“What cannot, is not, and will not be tolerated anywhere is infringement of songwriters’ and composers’ rights,” said John Phelan, director general of international music publishing trade association ICMP, in a statement. The AI Act, he says, will ensure “special attention for intellectual property rights” but further improvements to the legislation “are there to be won.”
More than six decades after their formation, Paul McCartney says the final-ever Beatles song is on its way thanks to the miracle of modern technology. Speaking to BBC Radio 4’s Today, Sir Paul said that he has been using artificial intelligence to “extricate” John Lennon’s voice from an old demo to complete the untitled track.
“We just finished it up and it’ll be released this year,” he said, of the untitled song that the BBC speculated could be a 1978 Lennon composition called “Now and Then.” The single was reportedly in the running to serve as a “reunion song” for the 1995 Anthology series, which included two new songs based on demos recorded by Lennon after the group split, 1995’s “Free As a Bird” and 1996’s “Real Love,” produced by ELO’s Jeff Lynne. Those tracks were the first “new” Beatles” releases in more than 25 years.
McCartney reportedly received the demo for the new track from Lennon’s widow, Yoko Ono, in 1994; the song was one of several on a cassette labelled “For Paul” that Lennon made shortly before his murder in 1980. The BBC reported that the tracks were “lo-fi and embryonic” and mostly recorded on a boombox by Lennon on a piano in his New York apartment.
The BBC reported that the living members of the band tried to record the “apologetic” love song “Now and Then” around the time of the Anthology release, but abandoned the sessions in short order. “It was one day — one afternoon, really — messing with it,” Lynne said. “The song had a chorus but is almost totally lacking in verses. We did the backing track, a rough go that we really didn’t finish.”
McCartney later said guitarist/singer George Harrison refused to work on “Now and Then,” saying the sound quality on Lennon’s vocals was “rubbish… George didn’t like it. The Beatles being a democracy, we didn’t do it.” The BBC reported that there were also reportedly technical issues with the original, due to some persistent “buzz” from the electrical circuits in Lennon’s apartment. The new version of the demo reportedly popped up on a bootleg CD in 2009, minus the background noise.
In a 2012 BBC documentary on Lynne, McCartney said, “that one’s still lingering around… so I’m going to nick in with Jeff and do it. Finish it, one of these days.” And while it is still unknown if that song is the one due out, the BBC reported that technical advances employed during the making of Peter Jackson’s Get Back Beatles documentary series — during which dialog editor Emile de la Rey trained computers to recognize the Beatles’ voices and separate them from background noise, including their own instruments — allowed the team to create “clean” audio. That same technology also allowed McCartney to sing a virtual duet with Lennon on his most recent tour.
“He [Jackson] was able to extricate John’s voice from a ropey little bit of cassette,” McCartney told Radio 4 in explaining how the tech used in the documentary helped him work on the “new” song. “We had John’s voice and a piano and he could separate them with AI. They tell the machine, ‘That’s the voice. This is a guitar. Lose the guitar.’ So when we came to make what will be the last Beatles’ record, it was a demo that John had [and] we were able to take John’s voice and get it pure through this AI. Then we can mix the record, as you would normally do. So it gives you some sort of leeway.”
At press time a release date for the Beatles track had not been announced.